What problem does it solve? Deep architectural analysis, root cause debugging, and implementation planning require focused reasoning that is hard to do alongside active coding. This Skill delegates that work to a dedicated architect subagent that analyzes specified files and returns structured strategic guidance without making code changes. ## Core Features & Use Cases - Architect Subagent: Spawn an analysis-focused subagent with a defined task, relevant files, and a responsibility mode. - Three Responsibility Modes: Use "plan" for implementation planning, "debug" for root cause analysis, and "evaluate_task" for reviewing completed work. - Git Context Integration: Include the current unstaged diff or a commit range (e.g., "HEAD~3..HEAD") so the architect understands recent changes. - Use Case: After building a major feature, ask the architect to evaluate the implementation against design goals, or debug why a session validation function rejects valid tokens. ## Quick Start Ask the architect to review the authentication module files and debug why token validation fails for valid sessions.